Tlie Study of Living Languages, [no. 4, new series. 
be particularly observed that it is not the bare knowledge of a word 
that is wanted, so that a person by more or less thought can recall 
it to his mind, it must, to be of any use in conversation, be per- 
fectly familiar, and it must have been repeated aloud so many times 
that the organs of speech have been thoroughly exercised in it, and 
that it may be pronounced both correctly and with the utmost free- 
dom. Nothing but multiplied repetitions of it aloud with continued 
reference to a correct standard^ and in connection with various other 
words can accomplish this. It must take some time thus to appro- 
priate new words, and especially the first 100 words of a language, 
but the process cannot possibly be hastened, h'h.'i on the contrary 
indefinitely delayed, by attempting thousands"of others before the 
first are secured. 
The number 100 is chosen for the first batch, as being about the 
smallest number that can enable one to make up some variety of 
short sentences, so that each word may be seen in a variety of si- 
tuations and in its various infiections ; and they are sufiicient to 
j)rovide for exercising the student in the first rudiments of grammar. 
3rd. To each of these batches of words a set of sentences is to 
be added ; these are all to be the commonest colloquial expressions. 
They should consist of a certain number written on each word in 
the batch, and contain no words that are not in it. It is essential 
that they should be written by Natives, and those who cannot speak 
English would be preferable, in order that there may be a security 
for the sentences being true Native expressions. It is also essen- 
tial that the writers should not be highly educated men, but ordi- 
nary intelligent men of the middle classes, otherwise the sentences 
would be almost sure to be full of fanciful things. The list of 
words should be given to several diff'erent persons in order to se- 
cure a good variety of expressions, and some selection should be 
made. The first set ought not to consist of less than 1,000 sen- 
tences, thqA is 10 to each word, in order that by means of this 
batch of words the first rudiments of grammar may be quite fa- 
miliar, and some considerable notion obtained of the general 
style of expression peculiar to the language, keeping throughout 
the grand object in view, which is to arrange that, so far as pos- 
